Verified65% confidenceFactExact time
Speculative decoding was formally proposed by Google in 2023, using a draft model with 1/10 to 1/50 the parameters of the target model to generate multiple candidate tokens at once, which are then verified in parallel by the large model.
3
Sources
65%
Confidence
Long-term
Relevance
7/22/2026
First Seen
Sources
Related Claims
UnverifiedSpeculative decoding was proposed by Google in 2022 and can achieve a 2–3x throughput improvement without sacrificing output quality80% similarUnverified推测性解码由Google Brain于2023年正式提出，核心洞察是GPU在验证一批Token时的计算成本与验证单个Token相近76% similarUnverified推测解码由Google DeepMind的Leviathan等人和陈等人于2023年独立提出，基于拒绝采样的验证算法确保输出分布与主模型完全等价75% similarVerified思维链（Chain-of-Thought）提示技术由Google Research于2022年提出74% similarUnverifiedGoogle's official documentation states that all parameters of the 26-billion-parameter MoE model are loaded to maintain inference speed, even though only 4 billion parameters are activated.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/588978API
curl https://kongchang.com/api/v1/knowledge/claims/588978MCP
get_claim(id=588978)